Getting at the truth about Bayh-Dole’s impact, Part 1

Last December Catherine Kirby, a student at Rice University, posted “The True Impact of the Bayh-Dole Act” at the McNair Institute for Entrepreneurship and Innovation. Her article got called out in a tweet by Daniel Garisto as an instance of the welter of conflicting claims about Bayh-Dole. While we may not hold a student writing a blog post at a university entrepreneurship center to the same standards as, say, an Ashley Stevens writing articles that make an implicit claim (at least) for being authoritative, Kirby’s work does provide evidence for what someone reviewing the spin on Bayh-Dole might reasonably surmise–that Bayh-Dole has had an important impact, and that impact has, at least for biotech, been generally positive, and that’s the “truth.”

Let’s work through Kirby’s discussion as a document that repeats the things anyone can find in a casual inspection of the upper reaches of a Google search on Bayh-Dole’s impact. It’s a clue perhaps that the primary literature questioning Bayh-Dole that shows up early in a Google search comes from authors in other countries debating whether Bayh-Dole has done what’s claimed, and whether what is claimed will do the same thing for their country. If creating fragmented ownership, institutionally licensed technology outside public oversight and without government intervention is the goal, then Bayh-Dole may be just the ticket. If there’s a strong wealthy class with a fixation on monopolies *and* a willingness to pay to acquire those monopolies, then one can even create a nice speculative market in patents on research findings. No matter whether anything actually arises from the use of the inventions that motivate the trade in such patent properties. The prospect of new patent monopolies alone is sufficient for those that have the monopolies to separate money from those that want the monopolies and potentially even more money from those that ignore the monopolies or reject them.

As for Bayh-Dole, keep in mind that the law makes its own metrics a government secret. Any account of the “true” impact of Bayh-Dole needs to consider the obvious: information about the deployment of university research inventions under patent–posi–has become a secret. That’s an impact of Bayh-Dole. Right there in the law. All we get is the positive bit, made to appear representative of the whole. The University of California recently reported that only 0.5% of its inventions (overall, not just subject inventions) resulted in commercial products (and with no information whether those commercial products actually had much market). Given that many of the top producing inventions in the UC system are plant varieties–strawberries, especially–one might even come to think that the emphasis on technology innovation is rather overblown.

If the 0.5% figure is anywhere near the mark, and if UC is at all representative of the state of licensing practice, then university commercialization rates have fallen from the 25% to 50% claimed for non-subject inventions in the years prior to Bayh-Dole by a factor of perhaps 100x. Even under the IPA program, universities reported only about a 5% commercialization rate for subject inventions. When pushed by a Congressional committee on their lack of effectiveness, the best answer the university representatives could make is that they might “get better” if they were given outright even more federally supported inventions to work with.

Here’s an opening bit from Kirby’s post:

Some credit the act as an engine of economic growth responsible for the emergence of the biotechnology industry. Critics say that the law decreased data sharing and basic research and increased health care costs. Others think that the act had little impact and that changes in university patenting were inevitable.

Not much to argue with there, other than that the number of folks writing in support of Bayh-Dole way outnumber those that are critical. Grassroots or the agenda of those who benefit from Bayh-Dole? You decide. If numbers indicate truth, then hey, all we would have to do is count. Not many university researchers champion Bayh-Dole. Not many small business owners. Not a whole lot of members of the general public going, wow, where would we be without Bayh-Dole? The majority of folks advocating Bayh-Dole are in organizations that benefit–university administrators, patent brokers, patent attorneys, companies that work in monopoly markets.

What’s odd is that there aren’t many critics pointing out that Bayh-Dole’s rate of practical application (broader than commercial products) appears to be 1/100th of the university outcomes prior to Bayh-Dole. The critics cannot even get at the core metrics of the law. Ah, that right there must be one of the successes of the Act–that it covers its own tracks so even academics can’t figure out what to examine.

Consider, though, what comes next:

University patenting would have increased regardless of the Bayh-Dole Act.

David Mowery shows that university patenting was already increasing before Bayh-Dole. That may have been the result of multiple motivations–first, there was the MIT-UC-Stanford move to establish campus-based IP offices with an express goal of generating additional research funding from IP positions. That move, in turn, appears to have resulted from a combination of dissatisfaction with Research Corporation and other patent management agents and with a certain administrative lust to make sure that any profits from inventions made in research hosted at one’s university should come back only to that university and not be spread around, as Research Corporation was wont to do, to the best research that it could find to support.

Was the motivation, then, to find ways to get more money for research? If so, why would patenting in the 1980s be seen as a need? Was it because the federal budget for university research, which had seen significant growth in the 1950s and 60s, had flattened in the late 1970s and 1980s, and the universities were looking for support for the researchers who had flooded into the universities as they expanded to swallow all that additional federal funding? That is, was patent licensing income sought to supplement federal funding to permit universities to continue to expand their research base, even with a decrease (or at least a decrease in the growth) of federal funding? In turn, was the need to grow a university’s research enterprise a matter of competing with other universities for what funding was available, and in so doing to enhance the luster of the university’s reputation?

There’s no question that university administrators have spent a great deal on shining up institutional reputations–game the rankings, issue glossy advertisements, write articles about economic impact. Research reputation in turn might recruit more faculty with grants, leading to more research expansion–construction, grants, donations–and supporting higher tuition and (for state schools) more legislative support (for research, for economic development, not for teaching). Perhaps then patent licensing income was thought to provide more resources to be even bigger in research, and hence move up in the rankings, and thus appear more excellent. Once it became clear that other than a handful of big hits, patenting was mostly a dead weight, administrators turned to counting patents as a virtue sign of the potential for economic impact and counting licenses as a further sign of that potential for impact, now so close that a license could be left out there as an all-but-commercialization fact.

I’ll point out one more possible use of an internal patent licensing operation–even a money-losing one. An internal patent operation at a public university may be funded by state allocations. Thus, the money the office loses is public money. But the money it generates–if and when it does–is free and clear of public claims. Royalty money, after paying inventors and costs, can be used as slush funds (well, not actually, according to Bayh-Dole’s standard patent rights clause–but no one bothers to enforce that clause). Basically, a public university patent office can be another laundering scheme to convert dedicated state funding into discretionary administrative funding. Even if all the office does is break even, it converts state money into discretionary money. That function is a common scheme in university side dealing, and good as gold for administrators looking for spare money to use as they will. Not that critics of Bayh-Dole will bother looking for something as administratively practical as money laundering as an impact of the law.

There’s a second point worth raising, then, about this sentence on university patenting. Would university inventor patenting have increased without Bayh-Dole? That is, before Bayh-Dole (and before the revived IPA program), university inventors would have to decide they wanted to apply for a patent, and then seek out assistance either from the university or more likely an affiliated research foundation. Universities might take a financial interest in such efforts rather than demand an ownership position. If a faculty inventor could find an agent willing to front the cost and do the patenting and licensing work, why take on that bother and expense in an internal office? Why demand to own and then review every invention anyone possibly thinks up, even the ones that the inventors don’t think is worth the effort to patent?

Thus, it is possible that Bayh-Dole has been used as a driver by university patent administrators to acquire more patents–on stuff that was never worth patenting as a pre-condition for use, development, or product. Further, where patenting might have been indicated, the patents would have been owned by the inventors or by affiliated foundations or directly by a company. The number of university patents could have remained low–indicating voluntary reporting, institutional selectivity, and an operating model that preferred a financial interest in upsides to an ownership interest in many downsides in an effort to increase the number of upsides.

Bayh-Dole, then, may well have shifted invention work towards patenting, and shifted patenting away from agents and companies to university administrations and in doing so moved research inventions *further* away from early adopters. With university patent positions (especially those set up to default to exclusive licenses=assignments), a research finding is withheld from all use pending a decision regarding which favorite monopolist might take an exclusive license. That monopolist might not ever show up, but in the meantime, everyone else has to sit on their hands, or take the risk of infringing, or work to design around the invention, or work to block the improvement pathways and applications for the invention with surrounding patents, setting up an eventual cross license (which rarely happens–universities don’t cross license) or more likely a counter claim if sued (which also is often useless, as universities and their failed startup-cum-trolls are non-practicing entities). Imagine that–a university patent position inducing industry to drift away and work to undermine the invention rather than use it. Imagine that–a university patent position attracting speculators looking to profit from the patent right (right to exclude all others, right to sue users) in favor of making the invention available to everyone.

For all of that, then, an increase in university patenting may indicate more money being allocated to patenting, less selectivity in what to patent, more demands to patent even what inventors don’t believe ought to be patented, and patenting what otherwise would have been patented and deployed without the need to pass through a university ownership position. Patent count goes up. Nothing beneficial in the activity other than that useless designated public money (to comply with research contracts, to create economic development) can be converted to glorious discretionary money (royalties)–and all one needs is one big patent hit every two decades for this little engine to pay off.

There’s absolutely nothing about an increase in university patenting that points to a beneficial impact of Bayh-Dole, nor that such an increase was inevitable.

Part 2 is here.

This entry was posted in Bayh-Dole, Metrics, Policy. Bookmark the permalink.

Leave a Reply